Methods and Challenges in Estimating Fish Stock Abundance

Foundational Concepts in Fish Stock Abundance

In the study of fisheries science, understanding the status of fish populations is essential for effective management. Several key terms define how these populations are measured and categorized. Fish Biomass refers to the total weight of all fish within the ocean or a specific area, while Fish abundance is the metric for the total number of individual fish. Stocks are also classified based on their interaction with human activity: Unexploited fish are those that are not currently being fished, whereas Exploited fish are those that are actively caught by fishing operations.

Estimating the abundance of fish is notoriously difficult. As noted by John Shepherd, counting fish is comparable to counting trees, with the added complications that the fish are invisible to the naked eye and are constantly in motion. This inherent difficulty necessitates specific scientific methods to estimate population sizes accurately, including catch per unit effort ($CPUE$), underwater visual census, and mark-recapture techniques.

The Rationale for Monitoring Fish Stocks

Determining the geographical limits and the abundance of fish stocks is a primary concern at the inception of any fishery. This information serves two critical stakeholders. For fishers, having an estimate of the stock size is necessary to justify the significant financial investment required for boats, fishing gear, and shore-based facilities. For managers, these data are vital to determine sustainable fishing levels. Specifically, managers must decide how much fish can be harvested from a stock without causing a serious decline in the population, ensuring the long-term viability of the resource.

Absolute vs. Relative Abundance

Datasets regarding fish abundance are generally divided into two categories. Absolute abundance represents the actual, precise number of fish located in a specific area or the total population of a fish stock. In contrast, Relative abundance measures the number of individuals in the same area compared to another point in time. Relative abundance serves as an index rather than a total count.

If a fisher catches an average of $10$ fish per day in the year $2020$ and subsequently catches an average of $8$ fish per day in $2021$, the logical inference is that the fish abundance in the year $2021$ has decreased by $20\%$. The most frequently utilized index for measuring relative abundance is known as Catch Per Unit Effort, or $CPUE$. This index is based on the fundamental assumption that changes in the $CPUE$ accurately reflect underlying changes in the abundance of the fish stock.

Methods and Units of Catch Per Unit Effort (CPUE)

$CPUE$ data can be recorded using various metrics depending on the specific fishery and gear type. Standard examples include the number or weight of fish caught per hook per hour, the number of lobsters caught per trap per day, or the weight of demersal fish caught per hour of trawling operations.

Data for $CPUE$ is sourced from two main channels: fishery-dependent data and fishery-independent data. Fishery-dependent data is obtained directly from commercial fishers, often through the use of fishing logbooks. In many jurisdictions, completing and submitting these logbooks is a legal requirement. However, the quality of this data is heavily influenced by the fishers' support of fishery objectives; fishers who are actively involved in management are theoretically more likely to provide accurate records. Fishery-independent data is collected by scientific or management agencies through structured surveys.

Case Study: Yellowfin Tuna and SubAntarctic Trawl Surveys

Evaluating the fishery-dependent $CPUE$ for yellowfin tuna in the eastern Atlantic provides a clear example of how catch and effort interact over time. The formula used for this calculation is $CPUE = \frac{\text{Catch}}{\text{Effort}}$. Data from this fishery between $1965$ and $1988$ shows various trends. For instance, in $1965$, the effort was recorded at $27.72042571$ days with a catch of $67.721 \times 10^3$ tonnes, resulting in a $CPUE$ of approximately $2.443$. By $1977$, the effort increased significantly to $191.259854$ days with a catch of $131.013 \times 10^3$ tonnes, which saw the $CPUE$ drop to $0.685$.

In fishery-independent contexts, such as the SubAntarctic trawl survey, data collection is often more sporadic. Records from $1972$ to $2016$ show large gaps where data was not available (marked as #N/A). For example, in $1992$, the catch was $138,600$ tonnes with an effort of $80,285$, yielding a $CPUE$ of approximately $1.7263$. By $2004$, a catch of $57,900$ tonnes with an effort of $14,318$ resulted in a higher $CPUE$ of $4.0439$. In $2015$, the catch was $96,900$ tonnes with an effort of $31,329$, resulting in a $CPUE$ of $3.093$.

Limitations and Pitfalls of Using CPUE

While $CPUE$ is a standard tool, there are significant risks in using it as a sole index for abundance. One major issue is that $CPUE$ can fluctuate widely due to factors unrelated to the fish population, such as the specific luck or skill of the fisher. This volatility makes it difficult to distinguish between environmental noise and actual changes in stock size.

Another significant pitfall involves the spatial distribution of the fish stock and the subsequent fishing effort. Fishers typically begin by targeting inshore fish stocks. As those stocks are depleted, fishers move further offshore to maintain high catch rates. In this scenario, the $CPUE$ remains high, which may lead to an overestimation of abundance. The data would suggest the stock is healthy because catch rates are stable, even though the inshore portion of the stock has been significantly depleted.

Comparative Analysis of Data Sources

Fishery-dependent data is characterized by its high quantity but potentially lower quality. Concerns exist regarding the accuracy of logbooks and the potential for bias, as commercial fishers naturally congregate in areas of high fish density. This effectively samples only a small, productive portion of the total stock. However, this data is very cheap to collect relative to scientific surveys.

Fishery-independent data is often of higher quality and more accurate because it covers a wider area of the stock's distribution and follows scientific protocols rather than commercial motivations. The primary drawback is the cost; research vessels are the most expensive type of scientific equipment. Consequently, the quantity of fishery-independent data is usually small compared to the vast amount of data generated by commercial fishing operations.